Contents lists available at SciVerse ScienceDirect
Engineering Geology
j ourna l homepage: www.e lsev ie r .com/ locate /enggeo
Characterization of aeolian sands from Indian desert
G.P. Padmakumar, K. Srinivas, K.V. Uday, K.R. Iyer, Pankaj Pathak, S.M. Keshava, D.N. Singh ⁎Department of Civil Engineering, Indian Institute of Technology Bombay, Powai, Mumbai-400076, India
Article history:Received 7 November 2011Received in revised form 10 April 2012Accepted 12 April 2012Available online 26 April 2012
Keywords:Aeolian sandThe Great Indian Thar desertCharacterizationConstruction materialIndustrial applications
Aeolian sands, from Sam, Jaisalmer, Rajasthan, India, belong to the Great Indian Thar desert and, primarily,attract tourists from India and all over the world. Though, certain studies were conducted on the geologicalorigin of these sand and deserts, results related to their physical, chemical, morphological, mineralogical,thermal, electrical and geotechnical characteristics are scanty. However, with an increase in demand of theland for infrastructure development (for residential, commercial and strategic facilities), for which their sta-bilization may be essential, and a need to utilize these (aeolian) sands as a construction material, particularlydue to acute scarcity of sand, their complete characterization becomes essential. With this in view, a detailedstudy was conducted to characterize these sands and details are presented in this paper. In general, thesesands are found to exhibit properties similar to aeolian sands from the Arabian Peninsula, Australia and China.Further, with an intention to utilize these sands in construction industry, especially as fine aggregate in concreteandmortar and designing thermal beds for buried conduits and electrical cables, the results have been comparedwith those for the Indian standard sands. Based on the chemical composition and chemical properties, and crush-ing strength of aeolian sands, their utilization in concrete and other construction materials (viz., bricks, buildingblocks, paver blocks etc.) appears to bequite promising. Also, the collapse potential of these sands has been foundto be quite lowwhile its angle of internal friction is quite high. These parameters suggest that aeolian sands fromthe Great Indian Thar desert can also be used for various engineering applications.
Sand dunes (refer to Figure 1) at Sam, Jaisalmer, India, are a partof the Great Indian Thar desert and are located near the internationalborder between India and Pakistan. This place is located within arectangle lying between 26° 4′–28° 23′ North parallel and 69° 20′–72° 42′ East meridians, as depicted in Fig. 2. This region experiencesarid climate and is a ceaseless sea of sand dunes, of all shapes andsizes, some rising to a height of 46 to 50 m. In this region, the maxi-mum and minimum temperatures, during summers, range from 42to 25 °C, respectively. Regular dust storms during summers make itdifficult for the local populace and tourists to stay there comfortably.However, during winters these temperatures are about 24 and 6 °C,respectively. The average annual rainfall in this area is only about150 mm.
Sam attracts tourists from India and all over the world due to itssand dunes. Some studies have been reported in the literature regard-ing the geological origin of these sands and deserts (Subramaniamand Kesava, 1983; Bakliwal and Wadhawan, 2003). However, notmany studies have been conducted by earlier researchers to establish
the physical, chemical, morphological, mineralogical, thermal, electricaland geotechnical characteristics, which are essential for (a) infrastruc-ture development (for residential, commercial and strategic facilities),for which their stabilization may be essential and (b) utilization ofthese sands (read aeolian sands) as a construction material. In thiscontext, the importance of complete characterization of aeoliansands gets highlighted by the following studies.
Aeolian sands are, primarily, unconsolidated sediments formed bythe erosion, transportation and deposition of the materials of weath-ering from the sandy parent material by the wind in the arid environ-ment (Milton, 1967; Al-Sayari and Zolt, 1978; Watson, 1985; Zhu,1985; Wu, 2003). A vigorous and continuous deposition of thesesands results into the formation of sand dunes, which have peculiarcharacteristics of changing their location, length and height, depend-ing on the direction and intensity of the wind (Tsoar, 2001).
It has also been reported in the literature that the mineralogicaland geo-morphological characteristics of the aeolian sands indicatetheir origin, type and process of formation (Ahlbrandt, 1979; Peaseet al., 1999; Abu-Zeid et al., 2001; Howari et al., 2007). These sands con-tainmore than 80% of the particles that are in the size range of 0.05 to0.25 mm (Chen, 1992). Incidentally, the particle size characteristicsand the mineralogical composition of the aeolian sands from theBlayney region of Australia were determined by conducting radiometricsurvey (Bruce and Scott, 1998). Themorphological and physico-chemicalproperties of the aeolian sands of northern China,were studied byZhenghu et al. (2007).
σ applied stressδ deformationω frequency of the ACτ shear stressμ viscosity of soil solutionσc electrical conductivityσcr crushing strengthγd dry densityεh, εv horizontal and vertical strains, respectivelyσn normal stressAS aeolian sandcc, cu coefficient of curvature and uniformity coefficient,
respectivelyCP collapse potentialcp specific heatD10,D30,D60 size fraction finer than 10%, 30% and 60%, respectivelyemax, emin maximum and minimum void ratio, respectivelyEC electrical conductivity of the soil solutionFM fineness modulusGs specific gravityk dielectric constantL/S liquid to solid ratioLR lime reactivityP crushing loadQ heat input per unit lengthr′ heater wire of resistancerp radius of the thermal probeRT thermal resistivityS1, S2 Indian Standard SandsSSA specific surface areaU electrophoretic mobilityα thermal diffusivityγdmax maximum dry densityγdmin minimum dry densityξ zeta potential
Fig. 2. The Google map of the location.
39G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
Some studies have also been conducted to stabilize dune sand ofKuwait and the Arabian Peninsula, for the purpose of providing certainoil, civil and military engineering applications (Phillips and Willetts,1978; Al-Sanad and Bindra, 1984). Detailed laboratory investigations
Fig. 1. The sand dunes at Sam,
on the sands form the dunes of Kuwait, for establishing their geotechni-cal properties, were conducted by Al-Sanad et al. (1993). In addition,studies have been conducted by earlier researchers (Diouf et al., 1990;Lahalih and Ahmed, 1998; Santoni and Webster, 2001; Asi et al., 2002;Basha et al., 2005; Al-Khanbashi and Abdalla, 2006; Han et al., 2007;Homauoni and Yasrobi, 2011) to establish the potential of various stabi-lizing agents (viz., acids, enzymes, lignosulfonates, tree resins andpolymers) for improving the engineering properties (i.e., stabilization)of the aeolian sands.
As mentioned earlier, such studies on the aeolian sands of Indiaappear to be scanty and hence could not be located in the existing lit-erature. Also, it has been felt that a comprehensive study includingphysical, chemical, morphological, mineralogical, thermal, electricaland geotechnical characteristics of these sands will open up the ave-nues for their proper utilization for various civil engineering applica-tions (infrastructure development, which may demand stabilizationof these sands and application as a construction material). With thisin view, complete characterization of aeolian sands, collected fromSam, Jaisalmer, Rajasthan, India, was conducted and details are pre-sented in this paper.
2. Geological implications
The Great Indian Thar Desert which is around 640 km long and480 km wide is surrounded by the Aravalli hills, the Rann of Kutch,
the Indus valley and the Punjab planes. The prolonged and excessivedegree of aridity combined with the sand drifting action is the trait ofthe Jaisalmer region which is a part of the Thar Desert. The climateshift, from warm humid to hot arid and semi-arid with periodic wetphases, resulted in complex and heterogeneous landforms of theJaisalmer Basin (Swain et al., 1983; Ramakrishnan and Tiwari, 1999).Consequently, the Quaternary sediment formations of the region wereidentified through three key cyclic processes commencing with fluvial,passing through lacustrine and ending in aeolian environment (Singhviand Kar, 2004). The terrain dominantly consists of parabolic and coa-lesced parabolic sand dunes with 60–90% fine non-calcareous sand(grains of size 0.18 mm are superseding) and 2–10% of silt-clay whichare devoid of vegetation (Subramaniam and Kesava, 1983).
Further it has been predicted that as a result of global warmingand associated climate change the desert zone will be subjected toeven higher temperatures and augmented aridity. Thus the degreeof hydrological and geomorphological processes shoots up which inturn leads to further extension of the desert land and thereby causeschanges in land use pattern as well as land degradation (Bakliwal andWadhawan, 2003).
3. Experimental investigations
Sampling of the aeolian sands was done over an area extendingabout 5 km. Five samples were collected from the locations 1 kmapart. Later, these samples were mixed together to form a representa-tive sample (denoted as AS). In addition, samples of Indian StandardSand (designated as S1 and S2) were also used in this study, as abenchmark material. Details of various tests conducted on thesesands, for establishing physical, chemical, mineralogical, morphologi-cal, electrical, thermal and geotechnical characteristics, are presentedin the following.
3.1. Physical characterization
3.1.1. Specific gravityThe specific gravity, Gs, of the sample was determined by using an
Ultra Pycnometer, (Quantachrome, USA) which utilizes helium gas asthe displacing fluid (as per the guidelines provided by ASTM D 5550).12 tests were conducted and the range of Gs obtained from these testsis listed in Table 1. Incidentally, the obtained Gs value, for the sand AS,is found to be much higher (ranging between 2.71 and 2.87) than thevalues reported for the dune sands of Arabian Peninsula (Al-Sanad etal., 1993). It should be noted here that for Indian Standard Sands(denoted as S1 and S2), the value of Gs ranges between 2.72 and 2.77and 2.85 and 2.88, respectively (refer to the values reported in paren-thesis in Table 1). Due to high Gs, the aeolian sands from Indian desertappear to be an excellent fill material, similar to their counterparts:sands S1 and S2, provided they could be transported in a suitable man-ner (may be in a slurried form) and placed at the construction site.However, this calls for a comprehensive study on rheological propertiesof these sands when they are converted to slurry by mixing with wateror a suitable solvent.
3.1.2. Void ratio and porosityThemaximum andminimum void ratios, emax and emin, correspond-
ing tominimum dry density (γdmin) and maximum dry density (γdmax)of the sample were determined as per the guidelines provided by ASTMD 4254-93. A Perspex mold of 50 mm inner diameter and 100 mmheight was employed for this purpose. To obtain the maximum voidratio, the sample was filled in the mold and its weight was determined.However, for determining the minimum void ratio, the mold with thesample was subjected to vibration for a period of 15 min, by installingit on a shake table. From the results presented in Table 1, it can be ob-served that sand AS exhibits very high value of void ratio as com-pared to its counterparts: sands S1 and S2, which have been tested
by the authors and Bartake (2006). It can also be observed that,though γdmin for the sand AS and dune sand of the Arabian Peninsula(Al-Sanad et al., 1993) match very well, γdmax for sand AS is on a muchhigher side as reported in Table 1.
3.1.3. Gradational characteristicsThe particle-size distribution characteristics of the sample were
determined by conducting the sieve analysis, as per ASTM D 422-63,and results are presented in Fig. 3. It was observed that no substantialamount of the sample passes through the 125 micron sieve and henceultra sieves (Shanthakumar et al., 2010) could not be employed forestablishing the particle-size distribution characteristics of the aeolian
41G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
sands. Subsequently,D10,D30 and D60 corresponding to the size fractionfiner than 10%, 30% and 60%, respectively, the uniformity coefficient, cu,and the coefficient of curvature, cc, were determined and the results arelisted in Table 1. Based on the USCS (ASTM D 2487-93, 1994), this sandcan be classified as poorly-graded sand (SP). The size range for sand AShas been found to be falling in the range (0.075 to 0.425 mm) reportedfor the sands from Arabian Peninsula and Australia, as reported bythe earlier researchers (Al-Sanad et al., 1993; Bruce and Scott, 1998;Howari et al., 2007). For the sake of better understanding and quickcomparison, the results for sands S1 and S2 along with the sands ofGrades I to IV (IS: 383, 1970) are also superimposed in Fig. 3. From thisfigure, it can be observed that the sand AS is finest among all and its fin-est fractions match very well with those of the Grade IV sand. For thesake of completeness, cu and cc values for sands of Arabian Peninsula(Al-Sanad et al., 1993), S1 and S2 are also presented in Table 1.
3.1.4. Specific surface areaThe specific surface area, SSA, of the sample was determined by
Blaines' air permeability apparatus, by following the methodologyprescribed by ASTM C 204, and it is depicted as SSA (Table 1). Portlandcement was used as the standard referencematerial for the test and thespecific surface area of the sample (in m2/g) was computed by usingEq. (1). The value obtained from this test is presented in Table 1.
SSA ¼ Sc 1−ecð Þffiffiffiffiffie3
p ffiffiffiT
pffiffiffiffiffie3c
q ffiffiffiffiffiTc
p1−eð Þ
ð1Þ
Where Sc is the specific surface area of the Portland cement(0.346 m2/g), e and ec are the void ratios of the sand sample and cement(=0.5), respectively, Tc is the time interval of manometer drop for
cement (77.18 s) and T is the time interval of manometer drop forthe sand sample.
3.2. Chemical characterization
3.2.1. X-ray fluorescence (XRF) analysisChemical composition of the sample, in its oxide form, was deter-
mined by employing an X-ray fluorescence instrument (Philips 1410,Holland). The sample was prepared by mixing thoroughly 4 g of thefinely grounded soil and 1.0 g microcrystalline cellulose with isopropylalcohol. The mixture was then kept below an infrared lamp for slowdrying. A small aluminum dish (inner diameter 33 mm and height12 mm) was taken and the two-third of the dish was filled with themixture of 70% methyl-cellulose, 30% paraffin wax and the remainingwith the dried sample. The sample thus obtained was subsequentlycompressed with the help of a hydraulic jack (by applying a 15 tonload) to form a pallet. Afterwards, the sample was mounted on themonochromatic sample holder of the XRF setup for determining itschemical composition and results are presented in Table 2. It can alsobe observed that the percentages of the major oxides for the sand ASare similar to the dune sands from other regions (Al-Sanad et al.,1993; Bruce and Scott, 1998). However, the silica content of sands S1and S2 is substantially higher (about 11 to 13%) as compared to thesand AS.
3.2.2. pH, electrical conductivity and total dissolved solidsThe pH of the sample was determined as per IS:2720 (Part XXVI)-
1987, by using a digital pH meter (Elico Private Ltd. Make, Model L1-120). The pH was established corresponding to liquid to solid ratio,L/S, varying between 2 and 10. The pH meter was calibrated by usingdifferent standard buffer solutions (pH=4, 7 and 9.2) prior to eachmeasurement. 30 g sample was mixed with distilled water and the re-sultant suspension was stirred thoroughly. The solution was allowedto stand for 1 hour and the stirring was done intermittently. The pHvalue of the solution was recorded corresponding to 20 °C at regulartime intervals. The same instrument, with automatic temperature com-pensator (corresponding to 25 °C) was employed for measuring theelectrical conductivity, EC, of the soil solution and total dissolvedsolids, TDS. The measured values can be considered as the back-ground values of the sample. The results of pH, EC and TDS with differ-ent liquid to solid ratios (L/S), for different sands, are presented inTable 3. When the results of pH are plotted as depicted in Fig. 4, it canbe noted that sand AS is more basic in nature as compared to sandsS1 and S2. However, the overall variation of pH with L/S is observedto be random. Due to the basic nature of the aeolian sand from India,they are best suited as aggregate in lean concrete (i.e. for plasteringwork, creation of footpaths and pathways), which also gets substantiat-ed by the fact that their fineness modulus, FM, is 1.20 (refer to Table 1).Incidentally, pH of the sand AS is found to be in agreement with the re-sults reported by Zhenghu et al. (2007), for the aeolian sands of
6 7 8 9 10
.058 8.719 8.823 8.880 8.609 7.909
.920 6.910 6.765 6.572 6.338 6.043
.382 6.817 6.256 6.735 6.958 6.795
.46 71.12 68.95 64.80 55.05 60.18
.09 44.31 38.94 38.15 31.73 31.09
.79 47.98 34.03 32.31 33.35 31.21
.34 35.69 34.59 32.54 27.51 30.44
.08 22.45 19.62 19.45 15.98 15.88
.77 24.23 17.35 16.94 16.40 15.82
0 1 2 3 4 5 6 7 8 9 10 11 120
1
2
3
4
5
6
7
8
9
10p
H
L/S
Fig. 4. The variation of pH with liquid to solid ratio for different sands.
0 2 4 6 8 10 120
20
40
60
80
100
TD
S(p
pm
)
L/S
Fig. 6. The variation of total dissolved solids with liquid to solid ratio for different sands.
northern China (=7.76 to 8.57). Furthermore, EC and TDS values of thesands AS, S1 and S2 are found to decrease with an increase in L/S, asdepicted in Figs. 5 and 6, respectively. Unfortunately, due to lack ofsuch data for the dune sands from other regions, validation of theobtained results could not be done.
3.2.3. Zeta potentialThe influence of pore fluid on particle-to-particle interaction in
the sample can be explored by the change in the surface charge po-tential which is indirectly defined as the zeta potential, ξ (Sparks, 1986;West and Stewart, 1995; Vane and Zhang, 1997; Yukselen and Kaya,2003).
The ξ of the sample was determined by employing an automatedelectrophoresis instrument (Zeta PALS, BIC, USA). This instrumentworks on the principle of ‘light scattering technique’, that determinesthe electrophoreticmobility,U (inm/s),which is the velocity of a particlein the solution produced by an external electric field of certain strength.U can be used to compute ξ by employing the Helmoltz–Smoluchowskitheory, which can be represented as (Kaya et al., 2003):
ξ ¼ 4π⋅m⋅U=kð Þ ð2Þ
where μ and k are the viscosity and dielectric constant of the soil solution,respectively, and U is the electrophoretic mobility. Measurements wereconducted on 1.5 ml solution, by maintaining the L/S equal to 10 of thesolution corresponding to 25 °C. The zeta potential of the sand AS
0 2 4 6 8 10 120
20
40
60
80
100
120
140
160
180
EC
(µ
s/cm
)
L/S
Fig. 5. The variation of electrical conductivity of the solution with liquid to solid ratiofor different sands.
corresponding to L/S=10 is obtained as −11.22 mV. However, thezeta potential values of the sands S1 and S2 for the same L/S ratio havebeen found to be−25.54 mV and −22.64 mV, respectively.
3.2.4. Lime reactivityIn order to ascertain suitability of the sand AS, as a construction ma-
terial, its lime reactivity, LR, was determined by following the method-ology proposed byDalinaidu et al. (2007). Thismethodology is based onthe principle that the loss in electrical conductivity, EC, of the lime solu-tion is due to the adsorption and fixation of the Ca2+ on the surface andmatrix of thematerial. This mechanism is responsible for the formationof hydration gel and results in development of strength when free limeis in contact with the material.
125 mg of analytical grade calcium hydroxide was dissolved in100 ml of distilled water and the EC value (in mS/cm) of the solutionwas measured by using the digital pH meter corresponding to 25 °C.The sand sample was then mixed with the calcium hydroxide solutionto the L/S ratio of 50, stirred well for 45 min and the EC value of theresultant supernatant solution was determined. Then the LR valuein MPa of the sand was determined by using Eq. (3).
LR ¼ 0:25� ΔEC−2:5 ð3Þ
where ΔEC (in mS/m) is the change in electrical conductivity of thecalcium hydroxide solution when it interacts with the sand sample.
0 10 20 30 40 50 60 70 80 90 100 110 120 130
0
5000
10000
15000
20000
Inte
nsi
ty
2θ
Fig. 7. X-ray diffractograms of different sands.
43G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
As per ASTMC593, a pozzolanicmaterial is best suited for cement andconcrete if its LR≥4MPa, which corresponds to ΔEC≥26 mS/m. For thesand AS, the LR value is obtained as 32.8 MPa, which is an indicationof its suitability in concrete and mortar. The LR values of the sandsS1 and S2 have been found to be 20.2 and 24.0 MPa, respectively.
3.3. Mineralogical characterization
3.3.1. X-ray diffraction (XRD) analysisThe mineralogical composition of the sand samples was deter-
mined by employing an X-ray diffraction (XRD) spectrometer (Philips2404, Holland)which utilizes a graphitemonochromator and Cu-Kα ra-diation. The sampleswere scanned from2θ ranging from5° to 120°. Thepresence of various minerals in the samples was identified with thehelp of the data files developed by the JCPDS (Joint Committee onPowder Diffraction Standards, 1994). The X-ray diffractogram, referto Fig. 7, indicates that aeolian sands are basically crystalline. With the
a
Fig. 8. Laser micrographs of the grains o
help of JCPDS data files, the presence of quartz, plagioclase, rutile,mica, calcite and fluoropargasite minerals have been confirmed insand AS. However, standard sands S1 and S2 consist of the mineralquartz, predominantly.
3.4. Morphological characterization
3.4.1. Laser microscopy analysisThe morphology (i.e., shape) of the grains of the sands was studied
by employing laser confocalmicroscopy. This technique,which employslaser-scanning, produces extremely high quality 3D images, as depictedin Fig. 8, of the grainswith the help of a lasermicroscope (LextOLS 4000,Olympus, Japan). From these images the average diameters of the sandgrains were computed and their values are found to vary between 0.16and 0.43 mm, 0.58 and 1.76 mmand 1.31 and 2.36 mm, for sands AS, S1and S2, respectively.
3.4.2. Scanning electron microscopy (SEM)In order to distinguish the environmental characteristics fromwhich
the sand grains were derived and to categorize the various surfacefeatures of the grains, scanning electron micrographs of the sandsamples were obtained by employing Quanta make 200 ESEM.
a b
c d
e f
Fig. 9. Scanning electron micrographs of th
Observations were carried out on a representative sample of grainsplaced in a sample holder, coatedwith platinum to avoid charge effectsand the images were captured at various magnifications, as shown inFig. 9. The micrographs at lower magnifications reveal the presenceof somewhat rounded, angular edged with flat surface, elongated
e sand (a, b) AS, (c, d) S1 and (e, f) S2.
45G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
and flaky grains for the sand AS (Figure 9a), and almost roundedsolid grains of varying surface irregularities for sands S1 (Figure 9c)and S2 (Figure 9e). As these sands consist of quartz, they exhibitcommon surface features like flat cleavage plates, upturned platesand conchoidal breakage patterns at higher magnifications (refer toFigure 9b, d, f) (Krinsley and Doornkamp, 1973).
The aeolian sand grains show a smooth rolling topography due tothe solution and precipitation taking place along with etching in theaeolian environment. Large grains (i.e. >200 μm in size) of the sandAS, exhibit upturned plates on the fractured portions, formed withwind abrasion which is responsible for knocking off the corners andthus the grains are almost rounded in shape. However, in smallsand grains (i.e., b200 μm in size), the flat upper and lower cleavagefaces and irregular upturned plates are noticed (Figure 9b). But theformation of conchoidal (i.e., irregular concave or convex curvatures)chips are observed on the surface of the grains owing to the excessivewind abrasion which in turn has reduced the grain roundness. Somereasonably angular grains having flat upper and lower surfaces, withirregular grain ends including conchoidal breakage features, havebeen noticed (Figure 9a) in the aeolian sands (i.e., sand AS). Edgesare also present in the grains and the cleavage plates are seen to belightly covered by a layer of silica. Single or multiple rounded, dish-shaped concavities are seen on the surface of the aeolian sand grains,which is an indication of the hot desert environment (Krinsley andDoornkamp, 1973).
The grains of the sands S1 and S2 are formed in the subaqueousenvironment which have been transported, abraded and deposited bywater forms. All the grains are having irregular rounded edges withirregular indentations at the surface as a consequence of mechanicalbreakage. Smooth surfaces are also seen especially in grains of sand S1as a result of rapid precipitation of silica (Figure 9d). In case of thesand S2, predominantly rough surfaceswith upturned cleavage surfaceshave been noticed (Figure 9f). The straight or curved grooves propagat-ing across the grain surfaces signify themechanical impact between thegrains under high energy conditions. The V-shaped patterns or depres-sions, which are observed irregularly across the grain structure of sandsS1 and S2, seem to be the diagnostic features of these sands from thesubaqueous environment (Krinsley and Doornkamp, 1973).
3.5. Electrical characterization
The electrical properties (viz. electrical conductivity, σc, and dielectricconstant, k) of the sand samples were determined by using an ‘Alpha-AHigh Performance FrequencyAnalyzer’ supplied byNovocontrol Technol-ogies, Germany, which operates over a frequency, ω, range of 3 μHz to
10-2 10-1 100 101 102 103 104 105 106 107 10810-3
10-2
10-1
100
101
102
103
σ c (
S/m
)
ω (Hz)
Fig. 10. The variation of electrical conductivity with frequency of AC for different densitiesof the sand AS.
40MHz. In order to calibrate the sample holder used for determiningthe electrical properties, distilled water was taken as the standard mate-rial. The sandwas filled in the sample holder, flushingwith the top end ofthe electrodes, corresponding todifferent drydensities,γd. This setupwasconnected to the two-wire high impedance test interface of the analyzerand electrical properties of the sample were recorded. The correlation ofthe frequency with σc and k for the sand AS is depicted in Figs. 10 and11, respectively. Trends depicted in these graphs indicate that forω>5 MHz, σ and k are practically independent of the γd of thesample, which is consistent with the results reported in the literature(Hanumantha et al., 2007). Further, k and σc of the sand AS were foundto be 3 and 10×10−5 S/m, respectively, forω>5MHz, by following themethodology presented in the literature (Shah and Singh, 2004). Inci-dentally, these values for sands S1 and S2 are reported to be 2.2 and1.66, and 16.5×10−7 S/m and 18.9×10−7 S/m respectively (Shah andSingh, 2004). A higher σc value of the aeolian sand, as compared to itscounterparts, can be substantiated by the presence of higher percent-ages of metal oxides viz., Na2O, K2O, Al2O3 and Fe2O3.
3.6. Thermal characterization
The thermal resistivity, RT, of the sand sample was found out byusing a thermal probe (Figure 12), which works on the principle of“Transient heat method” (Hooper and Lepper, 1950). The probe is ahollow copper tube of 60 mm long, 7 mm in outer diameter and0.5 mm wall thickness, which consists of a nichrome heater wire ofresistance, r′, 0.033 Ω/cm. The space between the heating elementand the copper tube is filled with thermal epoxy, which providesexcellent thermal conduction and acts as an electrically insulatedmaterial. The temperature of the probe can be determined with thehelp of a T-type thermocouple, which is attached to its surface. Itacts as a line heat source of input Q per unit length, of constant heatflux, in an infinite homogeneous medium initially at uniform tempera-ture which can bemeasured by knowing the applied voltage, Vp, acrossthe nichromewire (Gangadhara and Singh, 1999; Krishnaiah and Singh,2004, 2006). In order to ensure the proper functioning of the thermalprobe and for determining the thermal properties of the materials au-thentically, the probe was calibrated by using standard liquid glycerolwith a known thermal resistivity of 349 °C-cm/W. The standard liquidglycerol was filled up to the brim of a cylindrical perspex mold of90 mm inner diameter and 115 mm height. A top cover plate with acentral hole was fitted to the mold and the thermal probe was insertedin the glycerol through the hole. A constant voltage, V, was appliedwiththe help of a DC power supply unit and the temperature of the probewas data logged as a function of time. Temperature of the probe,when plotted against log of time yields the thermal response of the
100
101
102
103
104
105
106
k
10-2 10-1 100 101 102 103 104 105 106 107 108
ω (Hz)
Fig. 11. The variation of dielectric constant with frequency of AC for different densitiesof the sand AS.
glycerol. This procedure was repeated by applying different voltagesand by using Eq. (4), RT (in °C-cm/W) of the glycerol was computed.
RT ¼ s⋅Q4π
� �−1ð4Þ
where s is the slope of the straight line portion of the temperature, θ,versus log(time) plot in (°C) and Q is the heat input per unit length(=i2·r′), i is the current (in A) and r′ is the resistance of the nichromewire per unit length (in Ω/cm).
The thermal resistivity of the glycerol has been found to be 425,373, 275, 257, 221, 234, 215 and 229 (°C-cm/W) corresponding tothe input voltage of 0.2, 0.3, 0.4, 0.6, 0.8, 1.0, 1.2 and 1.4, respectively.The obtained resistivity values, when compared with the standardresistivity of the glycerol, indicate that the probe yields precise resultswhen the applied voltage is 0.3 V. As such, 0.3 V has been used asthe standard voltage for measuring thermal resistivity of the sandsamples.
For determining the thermal resistivity of the sand, the samplewas prepared in the same mold, which was used for calibration of
0.1 1 10 100 1000 10000
26
28
30
32
34
θ ((o C
)C
)
t (s)t (s)
s
Fig. 13. Thermal response of the sand AS (γd=1.46 g/cm3).
the thermal probe. A perspex plate (100 mm square and 10 mmthick) was used as a cover plate. This plate has a hole in the centerthrough which the thermal probe can be inserted in the sand.
0.3 V, which was identified as the optimum voltage from the cali-bration exercise, was applied with the help of a constant voltage DCpower supply unit and the temperature of the probe was data loggedfor about 10 min. The thermal response of the sample was obtainedby plotting the temperature of the probe against log of time, asdepicted in Fig. 13. This procedure was repeated for sands AS, S1and S2, by varying the dry density of the samples, and the corre-sponding RT values were computed and the same are presented inTable 4. Further, it can be observed from Fig. 14 that RT decreaseswith an increase in γd of the sample, for all the sands. This can be at-tributed to the fact that an increased density of the sample results in abetter grain to grain contact, displacing the air present between thegrains. The average values of thermal diffusivity, α, and specificheat, cp, were computed to be equal to 2.1×10−2 m2/s and 10.3 J/°C-g,respectively, by employing Eqs. (5) and (6).
α ¼ r2p=2:246⋅t0� �
ð5Þ
cp ¼ RT⋅γd⋅αð Þ ð6Þ
1.3 1.4 1.5 1.6 1.70
50
100
150
200
250
300
350
400
RT
T
(o C-c
m/W
)C
-cm
/W)
γd (g/cc) (g/cc)
Fig. 14. The variation of thermal resistivity of the different sands with dry density.
1000
10000
(kg)
(kg)
47G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
where rp is the radius of the thermal probe and t0 is the intercept onthe time axis of the θ versus log of time plot (Krishnaiah and Singh,2004, 2006).
It must be appreciated that thermal properties of the Indian aeoliansands are similar to Indian standard sands and hence their applicationfor designing fluidized thermal beds (Kolay and Singh, 2002) for theburied conduits and electrical cables will be quite appropriate.
-1 0 1 2 3 4 5 6 7 810
100
P
P
δ (mm) (mm)
Fig. 16. The load–deformation characteristics of different sands.
3.7. Geotechnical characterization
3.7.1. Crushing strengthThe crushing strength,σcr, of the samplewas determined by employ-
ing a 5 ton capacity servo hydraulic universal testingmachine, UTM. Thetest setup developed by Bartake and Singh (2005, 2007), depicted inFig. 15, was employed for this purpose. This setup consists of a stainlesssteel mold of inner diameter 33 mm, outer diameter 51 mm and height110 mm. The sample was filled to an aspect ratio (ratio of height of thesample to its diameter) of 1 as suggested by Bartake and Singh (2007) soas to achieve a density of 1.7 g/cm3. Two stainless steel pistons (top andbottom) of 32 mm diameter were used for crushing the sample. Silicongrease was applied on these pistons to minimize the side friction. Acollar is provided at the bottom of the mold on which two removablestainless steel clips of thickness 10 mm can be fixed. These clips wereremoved before commencing the test to achieve loading of the sam-ple from both ends. The tests were performed at a strain rate of1.25 mm/min (Karner et al., 2005). A 5000 kg load cell, attached toa digital read out unit was employed for recording the load, P, trans-mitted to the sample. The transmitted load was recorded at regularintervals until the load reaches 4500 kg. The resultant load versusdeformation (P vs. δ) characteristics were plotted, refer to Fig. 16,and σcr (=129 kg/cm2) was obtained following the methodologyexplained by Bartake and Singh (2005, 2007). The crushing strengthof sand AS is found to be almost equal to that of the standard sand S1and twice that of the sand S2. This indicates that aeolian sands fromIndia are extremely hard, which also gets substantiated by theirhigher Gs values. Incidentally, σcr of the sand AS lies in between thestrength of Grade II and Grade III sands (112.62 kg/cm2 and 150 kg/cm2, as reported by Bartake, 2006). Hence, further investigations should
Loading
Top Piston
SS
Collar
Bottom
Removable SS Clips
Fig. 15. The test setup used for determining crushing strength of different sands.
be conducted to establish proper utilization of the sand AS, as construc-tion material (viz., fine aggregate in concrete, artificial sands by provid-ing it an appropriate treatment). Such studies are being conducted bythe research group at IIT Bombay at this moment.
3.7.2. Shearing responseThe shearing response of the sand samples was determined by
using a direct shear box (100 mm×100 mm×42.5 mm) apparatus(HM-2560A Direct/Residual shear instrument, manufactured byHumboldt, USA). This is a computer controlled device and the samplewas sheared at a strain rate of 1.25 mm/min. Tests were conducted atthree different unit weights (γd=1.70, 1.75 and 1.80 g/cm3) of thesamples by subjecting them to different normal stress, σn, (=50, 100and 150 kPa), which was created by employing an air compressor.With the help of 2 LVDTs, connected to a data acquisition system, thehorizontal and vertical strains (εh and εv, respectively) and the shearstress (τ) were recorded. The stress–strain relationships, as depictedin Fig. 17, are noticed to be almost similar for all the sands. The angleof internal friction angle of the aeolian sands has been found to beequal to 40°.
3.7.3. Collapse potentialThe magnitude of one-dimensional collapse, due to inundation of
the dry sample,was determined. For this purpose, the collapse potential(CP) was determined by employing Eq. (7), as per ASTM D 5333.
CP ¼ e0−efð Þ= 1þ e0ð Þ½ � ð7Þ
where e0 and ef are void ratios of the sample in its dry state and afterinundation, respectively.
The variation of CP with the applied stress, σ, for the sand AS isdepicted in Fig. 18. From this figure it can be observed that the collapsepotential decreases as the normal pressure increases. The CP for sandASis found to be less than unity and hence it is not susceptible to collapseas per the recommendations by earlier researchers (Abelev, 1948;Denisov, 1963; Jennings and Knight, 1975).
4. Engineering geological implications
This study, in which aeolian sands from the Great Indian Thar deserthave been characterized for their various characteristics, unveils theirpotential as a construction material and imparts hope in utilizing thearid zone landforms for infrastructure development. The resemblancein the basic characteristics of these sands with the desert sands ofArabian Peninsula and other continents substantiates a great similarity
in terms of geological origin, in general. Furthermore, a comparison ofengineering properties of these sands with the properties (available inthe literature) of their counterparts, from other desert regions, providesenough confidence in adopting the construction practices, being pur-sued in the respective regions for developmental activities. It has alsobeen demonstrated that as aeolian sands exhibit analogous traits ofthe Indian standard sands, their utilization in construction industryis feasible. This would definitely help in overcoming scarcity of theriver sand and hence in preventing exploitation of the river basinsfor sand dredging.
Incidentally, results of various tests carried out for determiningphysical characteristics of these sands (viz., the specific gravity, void
0 50 100 150 200 250 300 3500.0
0.2
0.4
0.6
0.8
1.0
CP
(%
)
σ (kPa)
Fig. 18. The variation of the collapse potential with the confining stress.
ratio and porosity, grain size and distribution characteristics and spe-cific surface area) when critically compared with the results of dunesands of various continents, indicates their aeolian origin in the aridenvironment. The specific gravity and gradational characteristics ofthese sands point out their compatibility with the Indian standardsand and hence substantiate their utilization in various civil engineeringrelated applications. The chemical composition of these sands alsomatches well with their counterparts and as these sands are basicin nature, they exhibit substantially high lime reactivity. Due tothis peculiar characteristics, application of these sands as fine aggre-gate in concrete and mortar should be explored. The mineralogy ofthese sands, as identified by X-ray diffraction analysis, specifies itsgeological formation and surface texture. Themorphological analysisthrough the scanning electron micrographs of different magnificationsconfirms aeolian origin of these sands as well. Thermal properties ofthese sands validate its utilization as an alternative to the standardsands in thermal beds for underground power cables. Furthermore,the engineering properties of these sands such as reasonably highcrushing strength, higher angle of internal friction and low collapsepotential makes it a suitable constructionmaterial and for infrastruc-ture development. Thus this study can be considered as a steppingstone for better exploitation of the Indian desert sands, and thezones in which they are deposited, in a scenario of further extensionof the desert land and associated land degradation as a consequenceof global warming and augmented aridity.
5. Concluding remarks
This study deals with physico-chemico-mineralogical, thermal,electrical and geotechnical characterization of aeolian sands fromIndian desert. These sands are found to exhibit properties similar to theaeolian sands from the Arabian Peninsula, Australia and China. With anintention to utilize these sands in construction industry (especially asfine aggregate in concrete and mortar and designing thermal beds forburied conduits and electrical cables) the results obtained have beencompared with the characteristics of Indian standard sands. It hasbeen observed that except for the gradational andmineralogical charac-teristics, the aeolian sands from Indian desert exhibit almost similarcharacteristics as that of Indian standard sands. Based on the chemicalcomposition and chemical properties, and crushing strength of aeoliansands, their utilization in concrete and other construction materials(viz., bricks, building blocks, paver blocks etc.) appears to be quitepromising. However, handling and transportation of the aeolian sandsfor these applications appear to be a quite challenging task, which de-mands a meticulous study on their rheological properties. The collapsepotential of the soil is quite low and its angle of internal friction is quitehigh and as such, laying foundations of buildings on these deposits,without any suitable ground improvement, should not be a problem.
References
Abelev, Y.M., 1948. Fundamentals of Design and Construction on Macroporous Soils.Stroivoenmorizdat, Moscow.
Abu-Zeid, M.M., Baghdady, A.R., El-Etr, H.A., 2001. Textural attributes, mineralogy andprovenance of sand dune fields in the greater Al-Ain area United Arab Emirates.Journal of Arid Environments 48 (4), 475–499.
Ahlbrandt, T.S., 1979. Textural parameters of aeolian deposits. In: McKee, E.D. (Ed.), AStudy of Global Sand Seas United States Geological Survey Professional Paper,1052, pp. 23–32.
Al-Khanbashi, A., Abdalla, Sh.W., 2006. Evaluation of three water borne polymersas stabilizers for sandy soil. Geotechnical and Geological Engineering 24 (6),1603–1625.
Al-Sanad, H.A., Bindra, S.P., 1984. Soil Mechanics for Road Engineers in Arabian Penin-sula. Jami'at al-Kuwayt, Kuwait.
49G.P. Padmakumar et al. / Engineering Geology 139–140 (2012) 38–49
ASTM C 204, 2005. Standard test method for fineness of Portland cement by air perme-ability apparatus. Annual Book of ASTM Standards. ASTM, West Conshohocken, PA,USA, p. 10.
ASTM C 593, 1995. Standard specifications for fly ash and other pozzolans for use withlime. Annual Book of ASTM Standards, vol. 04.01. ASTM, West Conshohocken, PA,USA, pp. 314–318.
ASTM D 2487-93, 1994. Standard classification of soils for engineering purposes (Uni-fied Soil Classification System). Annual Book of ASTM Standards, vol. 04.08. ASTMInternational, West Conshohocken, PA, USA.
ASTM D 422-63, 1994. Standard test method for particle size analysis of soils. AnnualBook of ASTM Standards, vol. 04.08. ASTM International, West Conshohocken,PA, USA.
ASTM D 4254-93, 1994. Test method for minimum index density and unit weight ofsoils and calculation of relative density. Annual Book of ASTM Standards, vol.04.08. ASTM International, West Conshohocken, PA, USA.
ASTM D 5333-92, 1993. Standard test method for measurement of collapse potential ofsoils. Annual Book of ASTM Standards, vol. 04.08. ASTM International, West Con-shohocken, PA, USA, pp. 343–345.
ASTM D 5550-00, 2000. Test method for specific gravity of soil solids by gas pycnometer.Annual Book of ASTM Standards, vol. 04.08. ASTM International, West Conshohocken,PA, USA.
Bakliwal, P.C., Wadhawan, S.K., 2003. Geological evolution of Thar desert in India—issues and prospects. Proceedings of the Indian National Science Academy 69A(2), 151–165.
Bartake, P.P., 2006. Instrumentation for determining shear wave velocity in sands. Ph.D.Thesis submitted to Department of Civil Engineering, Indian Institute of TechnologyBombay, Mumbai, India.
Bartake, P.P., Singh, D.N., 2005. Determination of crushing strength of cenospheres.Journal of ASTM International 2 (7), 1–9.
Bartake, P.P., Singh, D.N., 2007. A generalized methodology for determination of crush-ing strength of granular materials. Geotechnical and Geological Engineering 25,203–213.
Basha, E.A., Hashim, R., Mahmud, H.B., Muntohar, A.S., 2005. Stabilization of residual soilwith rice husk ash and cement. Construction and Building Materials 19, 448–453.
Bruce, L.D., Scott, K.M., 1998. Recognition of aeolian soils of the Blayney district, NSW:implications for mineral exploration. Journal of Geochemical Exploration 63 (3),237–251.
Chen, L.H., 1992. Suggestion on aeolian sandy soil taxonomy. In: Gong, Z. (Ed.), Discussionon Soil Taxonomic Classification in China. Academic Press, Beijing, pp. 152–157.
Dalinaidu, A., Das, B.B., Singh, D.N., 2007. Methodology for rapid determination ofpozzolanic activity of materials. Journal of ASTM International 11 (6), http://dx.doi.org/10.1520/JAI100343 (Published Online: 11 July 2007).
Denisov, N.Y., 1963. About the nature of high sensitivity of Quick clays. Osnovaniya,Fundamenty i Mekhanika Gruntov 5, 5–8.
Gangadhara, R.M.V.B.B., Singh, D.N., 1999. A generalized relationship to estimate thermalresistivity of soils. Canadian Geotechnical Journal 36, 767–773.
Han, Z., Wang, T., Dong, Z., Hu, Y., Yao, Z., 2007. Chemical stabilization of mobile dunefields along a highway in the Taklimakan desert of China. Journal of Arid Environ-ments 68 (2), 260–270.
Hanumantha, R.B., Bhat, AjazM., Singh, D.N., 2007. Application of impedance spectroscopyfor modeling flow of AC in soils. Geomechanics and Geoengineering: An InternationalJournal 2 (3), 197–206.
Homauoni, Z.J., Yasrobi, S.S., 2011. Stabilization of dune sand with poly (methyl meth-acrylate) and polyvinyl acetate using dry and wet processing. Geotechnical andGeological Engineering 29, 571–579.
Hooper, F.C., Lepper, F.R., 1950. Transient heat flow apparatus for the determination ofthermal conductivities. Heating, Piping, and Air Conditioning 22 (8), 120–129.
Howari, F.M., Baghdady, A., Goodell, P.C., 2007. Mineralogical and geomorphologicalcharacterization of sand dunes in the eastern part of United Arab Emirates usingorbital remote sensing integrated with field investigations. Geomorphology 83,67–81.
IS: 2720. Part XXVI, 1987. Determination of pH value of soil samples. Indian Standards,Methods of Tests for Soils, New Delhi, India.
IS: 383, 1970. Specification for coarse and fine aggregates from natural sources for con-crete, Indian Standards, New Delhi, India.
JCPDS - Powder Diffraction File 44 7354 -CD ROM, 1994. International Centre for Dif-fraction Data, Pennsylvania, USA.
Jennings, J.E., Knight, K., 1975. A guide to construction on or with materials exhibitingadditional settlement due to collapse of grain structure. Proceedings 6th AfricanConference on Soil Mechanics and Foundation Engineering, Durban, pp. 99–105.
Karner, S.L., Chester, J.S., Chester, F.M., Kronenberg, A.K., Hajash Jr., A., 2005. Laboratorydeformation of granular quartz sand: implications for the burial of elastic rocks.AAPG Bulletin 89 (5), 603–625.
Kaya, A., Oren, A.H., Yukselen, Y., 2003. Settling behavior and zeta potential of kaolinitein aqueous media. Proceedings of the Thirteenth International Offshore and PolarEngineering Conference Honolulu, Hawaii, USA, pp. 407–412.
Kolay, P.K., Singh, D.N., 2002. Application of coal ash in fluidized thermal beds. Journalof Materials in Civil Engineering, ASCE 14 (5), 441–444.
Krishnaiah, S., Singh, D.N., 2004. Centrifuge modeling of heat migration in soils. Inter-national Journal of Physical Modelling in Geotechnics 3, 38–46.
Krishnaiah, S., Singh, D.N., 2006. Determination of thermal properties of soils in a geo-technical centrifuge. Journal of Testing and Evaluation 34 (4), 1–8.
Lahalih, S.M., Ahmed, N., 1998. Effect of new soil stabilizers on the compressivestrength of dune sand. Construction and Building Materials 12 (6), 321–328.
Milton, D.I., 1967. Geology of the Arabian Peninsula: Kuwait. American Geological SurveyF-567F, 560.
Pease, P.P., Bierly, G.D., Tchakerian, V.P., Tindale, N.W., 1999. Mineralogical characteri-zation and transport pathways of dune sand using Landsat TM data. Wahiba SandSea, Sultanate of Oman. Geomorphology 29 (3–4), 235–249.
Phillips, C.J., Willetts, B.B., 1978. A review of selected literature on sand stabilization.Coastal Engineering 2, 133–147.
Ramakrishnan, D., Tiwari, K.C., 1999. Calcretic and ferricretic duricrusts of the TharDesert, India: their geotechnical appraisal as a road paving aggregate. EngineeringGeology 53, 13–22.
Santoni, R.L., Webster, S.L., 2001. Airfields and roads construction using fiber stabilizationof sands. Journal of Transportation Engineering 127 (2), 96–104.
Shah, P.H., Singh, D.N., 2004. A simple methodology for determining electrical conduc-tivity of soils. Journal of ASTM International 1 (5) (Published Online: 10 May 2004.11 Pages).
Shanthakumar, S., Singh, D.N., Phadke, R.C., 2010. Methodology for determiningparticle-size distribution characteristics of fly ashes. Journal of Materials in CivilEngineering 22 (5), 435–442.
Singhvi, A.K., Kar, Amal, 2004. The aeolian sedimentation record of the Thar desert.Proceedings of the Indian Academy of Sciences — Earth and Planetary Sciences113 (3), 371–401.
Sparks, D.L., 1986. Soil Physical Chemistry. CRC Press, Boca Raton.Subramaniam, A.R., Kesava, R.A.V.R., 1983. Dew fall in sand dune areas of India. Inter-
national Journal of Biometeorology 27 (3), 271–280.Swain, A.M., Kutzbach, J.E., Hastenrath, S., 1983. Estimates of Holocene precipitations
for Rajasthan, India based on pollen and lake level data. Quaternary Research 19,1–17.
Tsoar, H., 2001. Types of aeolian sand dunes and their formation. In: Balmforth, N.J.,Provenzale, A. (Eds.), Geomorphological Fluid Mechanics. Springer, Berlin Heidel-berg New York, pp. 403–429.
Vane, L.M., Zhang, G.M., 1997. Effects of aqueous phase properties on clay particle zetapotential and electro-osmotic permeability: implications for electrokinetic soil re-mediation processes. Journal of Hazardous Materials 55 (1–3), 1–22.
Watson, A., 1985. The control of windblown sand and moving dunes: review of themethods of sand control in deserts with observation from Saudi Arabia. Journalof Engineering Geology 18, 237–252.
West, L.J., Stewart, D.L., 1995. Effect of zeta potential on soil electrokinesis. Proceedingsof Geoenvironment 2000, ASCE Special Publication, pp. 1535–1549.
Wu, Z., 2003. Geomorphology of Wind-drift Sands and Their Controlled Engineering.Science Press, Beijing. 448 pp. (in Chinese).
Yukselen, Y., Kaya, A., 2003. Zeta potential of kaolinite in the presence of alkali, alkalineearth and hydrolysable metal ions. Journal of Water, Air and Soil Pollution 145 (1–4),155–168.
Zhenghu, D., Honglang, X., Zhibao, D., Gang, W., Darke, S., 2007. Morphological, physi-cal and chemical properties of aeolian sandy soils in northern China. Journal of AridEnvironments 68, 66–76.
Zhu, Z.D., 1985. Status and trend of desertification in northern China. Journal of DesertResearch 5 (3), 3–11.
